2024: The Year the GPT Wrapper Myth Proved Wrong

Summary

Highlights

• Startups are seeing unprecedented success in 2024, proving the potential for rapid revenue growth. 🚀
• The initial belief that big companies like OpenAI would monopolize AI has been proven wrong as startups innovate and excel. ✌️
• AI innovations aren't limited to big enterprises; open-source and multi-model fields are hotbeds for creative startups. 🔥
• The trend towards vertical AI has sparked a range of industry-specific applications. 🌐
• AI development is faster and cheaper than ever, allowing for more agility and less dependency on large-scale funding. 💡

Key Takeaways

• Startups are thriving in AI, debunking the monopoly myth of big tech 🔥
• Multi-model architectures are revolutionizing AI applications 🚀
• Vertical AI is creating diverse opportunities across industries 🌟
• Open source is a game changer for model development & deployment 👍
• AI's impact on entrepreneurship—faster growth and reduced costs 📈

Overview

Chapters

• 00:00 - 02:30: Introduction and Startup Success The chapter discusses the current startup environment where it is possible to start a company that can generate tens of millions of dollars in a short time span, such as 24 months. Initial skepticism existed about whether enterprise pilots and proof of concepts could generate real revenue. However, recent experiences have shown that these pilots have indeed led to substantial revenue, marking a significant shift in startup success dynamics.
• 02:30 - 07:30: Chatbot Ecosystem Evolution The chapter titled 'Chatbot Ecosystem Evolution' discusses the rapid developments and changes in the chatbot industry as of 2024. The speakers reflect on the early days and express a sense of relief at current advancements, acknowledging the fast-paced nature of the industry. The chapter also highlights the achievements of notable individuals like Gary, Jared Harge, and Diana, who have collectively founded companies valued at hundreds of billions, signaling significant growth and impact in the tech space.
• 07:30 - 15:00: AI Model Development and Open Source Impacts The chapter discusses the year 2024 as a pivotal moment for startups, especially in the context of AI development. There is a reflection on the launch of Chat GPT two years prior and the initial expectation that open AI would capture all the value. Specifically, the chapter recalls the announcement of the Chat GPT store.
• 15:00 - 22:30: Revenue Growth and Market Trends The chapter discusses initial perceptions about the AI landscape, particularly focused on OpenAI and its application, ChatGPT. It highlights a past belief that many AI applications were merely extensions of ChatGPT, and that the release of the ChatGPT App Store would dominate, leaving little room for startups. However, in retrospect, these predictions seem misguided. The chapter notes that the ChatGPT store did not become significant, leading to a reconsideration of what constitutes successful AI applications beyond ChatGPT itself.
• 22:30 - 30:00: Startup Stories and AI Innovations The chapter "Startup Stories and AI Innovations" discusses the emergence of consumer and enterprise applications in the AI domain. Several successful startups have been highlighted, including Perplexity and Glean, along with legal tech companies like Case Tech and Harvey. The discussion emphasizes the rapid pace of building successful startups in the AI sector, with the potential to generate significant revenue in a short span, sometimes within 24 months. The chapter portrays the current period as an exciting time for startup ventures, especially in AI innovations.
• 30:00 - 37:00: Trends in Vertical Applications and Voice AI The chapter discusses the trends in vertical applications and Voice AI, specifically how startups like Opus Clip can succeed without raising a large amount of venture capital. This trend is observed across the Y Combinator (YC) community, highlighting the potential for startups to thrive post the launch of GPT-powered tools. The emergence of companies like Anthropic and Claude is mentioned, along with the industry's shifting perceptions towards capital requirements.
• 37:00 - 44:00: Robotics and AI Integration The chapter discusses the initial belief that only large foundation model companies like Amazon, Facebook, or Google, which have substantial financial resources, could compete in AI. It was thought that smaller entities could not bring value through AI applications unless they were developed by these large companies. This notion was later proven false, suggesting that even new players can hold significant value in AI integration.
• 44:00 - 53:00: Challenges and Flops in Technology The chapter explores the surprising development in technology driven by open-source movements, particularly focusing on Meta's unexpected release of LLaMA. Initially perceived as an innovative open-source model, LLaMA was, in reality, 18 months behind OpenAI. The chapter delves into the sequence of events that compelled Meta to release LLaMA, highlighting how the leak of model weights and subsequent decisions led to broader adoption and development within the tech community. This narrative exemplifies the unpredictable nature of technological advancements and the challenges companies face in keeping up with innovation.
• 53:00 - 60:00: AI-led Coding Revolution The chapter discusses how companies like Vicuna and other Llama-related ventures emerged, particularly focusing on a YC-backed company that facilitated local Docker development with models running on devices. Initially, there was skepticism about these models gaining traction. However, during the summer of 2023 to 2024, a significant shift occurred, marking a pivotal moment in the AI-led coding revolution.
• 60:00 - 68:30: Impact on Startup Hiring and Productivity The chapter titled 'Impact on Startup Hiring and Productivity' discusses a pivotal moment when the top foundation model rankings were led by 'llama', surprising the community. This shift highlights the significance of choice in models. While the quality of the model remains important, having multiple options prevents monopoly pricing strategies, thus impacting decisions related to hiring and productivity in startups.
• 68:30 - 75:30: Regulations, Politics, and Emerging Opportunities This chapter explores the critical aspects of a business model beyond the initial design. It emphasizes the importance of various factors such as product quality, the ability to sell effectively, responsiveness to user feedback, and achieving customer retention (zero churn). These elements are highlighted as being more crucial than merely focusing on future gains through the model itself. A specific example from a year ago is referenced, involving startups that were developing model routers, essentially APIs to call specific models. The insights indicate a shift in priority towards practical application and adaptability in business strategies rather than theoretical potential.
• 75:30 - 81:30: The Return of In-Person Events The chapter discusses the initial motivation for reducing the cost of using chat models by distributing interactions across different models, rather than relying on one. Initially, it was believed that the cost of these technologies would approach zero, thus eliminating the need for model routing in application development. However, a year later, this belief has been proven false, as using a single model is not always the best solution. The evolution of this perspective indicates a dynamic change in the field over time.
• 81:30 - 86:00: Conclusion and Outlook The 'Conclusion and Outlook' chapter discusses the use of model routers as an advantageous entry point for building applications powered by large language models (LLMs). It highlights a trend where companies prefer not to be tied to a specific model, a shift observed from previous seasons' trends in the startup batches. This approach allows companies to leverage multiple models, enhancing their adaptability and innovation.

2024: The Year the GPT Wrapper Myth Proved Wrong Transcription

• 00:00 - 00:30 the wildest thing right now is you can start a company that can make tens of millions of dollars literally in 24 months and you can do it for potentially you know2 million5 million a year ago I remember many of the startups in the batch would get sort of Enterprise proof of Concepts or pilots in particular and there was a lot of cynicism around whether any of those Pilots would translate into real Revenue fast forward a year I think we have all firsthand experienced that these Pilots have turned into like real Revenue it's still
• 00:30 - 01:00 early days honestly like it you know we sort of breathe a sigh of relief right now in 2024 but it's anyone's game honestly like these things are moving so [Music] quickly welcome back to another episode of the light cone I'm Gary this is Jared Harge and Diana and collectively we funded companies worth hundreds of billions of dollars right at the beginning so
• 01:00 - 01:30 2024 what a year how are you feeling about this Harge pretty great I think this is the year that everything broke in favor of startups what I've been thinking about a lot recently is when chat gbt launched two years ago now the immediate consensus viw was all of the value would go to open Ai and very specifically do you all remember when they announced the gbt or the chat gbt store yeah like the I remember the
• 01:30 - 02:00 consensus was everything that was built on top of chat gbt was a gbt rapper and the App Store was just going to be released and Crush every single person trying to build an AI application and open AI would be a ginormous company but there' be no opportunity for startups sounds kind of ridiculous to say that now because um who even remembers the chat GPT store exactly the chat gvt store itself was a nothing Burger but like more importantly what are the big AI applications today like I'd say outside of chat gbt itself the breakout
• 02:00 - 02:30 consumer application is perplexity the breakout Enterprise application is probably glean maybe um in legal Tech you have case Tech you have Harvey pruma you have photo room like there's the point being there are many many applications that have been built not by open AI it's been a great time to build startups yeah the wildest thing right now is you can start a company that can make tens of millions of dollars uh literally in 24 months from zero
• 02:30 - 03:00 and uh you can do it for potentially you know2 million5 million that's sort of the story of one of these companies Opus clip which never had to raise a real series a and that's something that we sort of see across the YC Community as well yeah I think that's um a particularly important point that you can do it as a startup without ra raising tons of capital because post the gbt store launch I then remember uh anthropic and Claude emerged and the consensus view for a while was all of
• 03:00 - 03:30 the value is going to go to one of these Foundation model companies and that the only way you can compete in AI is to raise huge amounts of money um either because you've got venture capital or your Amazon or Facebook or Google um with tons of cash already but that if you weren't one of the big foundation models um there would be no value and the applications built on top of these things would either be built by the foundation model companies themselves or just not be that valuable again something that turned out to be completely not true right and in
• 03:30 - 04:00 particular what drove that is open source like the weird series of events where me like the weight's being leaked and like meta just like rolling Tor yeah I kind of forced the hand for uh meta to launch a llama which was funny and people thought oh it was just this cool open source model but it was 18 months behind open Ai and people started doing a lot of der work out of it is like
• 04:00 - 04:30 vicuna and all these other animals related to llamas that came out and it took the or llas one of the companies at YC as well that enabled people to do uh local kind of Docker development like models running on device it was pretty cool but people didn't think that um they were going to be able to catch up and the thing that changed from 2023 to 2024 is that during the summer it was a
• 04:30 - 05:00 turning point it was the first time that the top foundation model in all the rankings benchmarks was llama and that was a shock to the community yeah so it turns out Choice uh matters and choice means that it's not as much about the model I think the model still matters quite a lot um but once you have choice in model it means you can't have the sort of idea of Monopoly pricing and you have that model you're pedor also has
• 05:00 - 05:30 that model but all the other things seem to end up mattering a lot more which is product your ability to sell your ability to actually adjust to user feedback your ability to get to zero churn all of those suddenly become far more important than capturing a light cone of all future value through the model right a very specific way I I felt this is I remember a year ago working with startups in the batch that are essentially building model routers just like an API to call like a specific model and I remember the um a lot of the
• 05:30 - 06:00 motivation for that at the time was uh reducing cost it was like oh like you don't want to just like burn up all of your chat BT calls you want to spread them out across like various different models and the the the argument against that was just oh like the cost of all this stuff is going down to zero anyway like there's no value to be had in being like a model router and no one wants to build their applications with a model router they're all just going to call whatever is the best model I think fast forward a year like that's totally not true like from what I can tell the the
• 06:00 - 06:30 model router was actually a really great entry point into just building sort of a new stack for building llm powered apps and most of the the applications we're seeing I think they just don't want to be beholden to a specific model does that map with what you've seen yeah actually one of the things we've seen now in the fall batch that just presented at demo day which was one of the trends that shifted from Summer 24 and winter 24 was precisely what you're saying companies started to use multiple
• 06:30 - 07:00 models for the applications like the best one for Speed at some point because sometimes you need to parse a lot of the input very quickly it's fine if it's a bit more lossy and then you need the bigger model to handle the more complex task so a lot of companies in Fall 24 have this actually multiple model architecture to use the best one for the best task which is similar to the concept of the model router but it was uh the idea evolve instead of being being more of a routing it was more of a
• 07:00 - 07:30 orchestration I think a concrete example we gave uh a couple episodes ago was uh camper was a company you work with they use the fastest model for paring like PDFs and the more complex ones they use 01 and that's that's how it's done and other companies that's doing uh fraud detection they have this concept of a like a junior risk analyst where they just use like a fast and easy gbd4 mini and and then they use the bigger one
• 07:30 - 08:00 with like o1 or the other example is um I think cursor talks about it in their episode with uh Lex Freedman they also have this complex multi-architecture with multiple models and this is why it works well it's like they do one very specific for predicting what you're going to type next but one for understanding the whole code base so very different tasks so that's definitely happening now yeah the other thing that uh popped up for fall batch there's a company I'm working with um called variant and uh what they're
• 08:00 - 08:30 trying to do is take basically state-of-the-art open- Source llm models that can do Cod genen and then teach them Aesthetics so uh starting with icon generation and so they built this huge sort of posttraining workflow that should work on you know as the open- source models get smarter and better at Coen broadly they can just you know take the next version of that and then uh take their post training architecture and data set and then
• 08:30 - 09:00 basically teach a given model Aesthetics so what a certain thing is supposed to look like and uh not in a diffusion sort of way but actually at the SVG level and we think SVG will actually translate into all kinds of Aesthetics so it's an interesting approach and like one of the newer ones and that post training is a whole coherent way to sort of skip the whole idea that um all the value is like
• 09:00 - 09:30 acur into the model especially because of Open Source to your point the other thing I I've been having flashbacks to is a year ago I remember many of the startups in the batch would get sort of Enterprise proof of Concepts or pilots in particular and there was a lot of cynicism around whether any of those Pilots would translate into real Revenue um lots of parallels to crypto and how anytime there's some new interesting technology blockchain more specifically than crypto but anytime there's a new technology Enterprises always want to run pilots and P because it's someone's
• 09:30 - 10:00 job to like check off yeah we did the like hot new technology thing the chief Innovation officer must have his VI we've spoken about this one of one of our episodes I think and yeah fast forward a year I think we have all firsthand experienced that these Pilots have turned into like real revenue and if anything the startups in the YC badge now are going to sell into real Enterprises faster than they have before and are ramping up revenue and reaching Milestones like a million dollars AR faster than I've suddenly ever seen yeah
• 10:00 - 10:30 the fall batch just did this actually again which is actually the first time I think we noticed it was actually the summer batch of this year and uh one of the funnier things that we realized was do you remember when Paul Graham uh would tell us how fast you needed to grow during the YC batch 10% a week 10% a week and uh the wild thing is in aggregate uh across both you know summer and fall batches that's what those batches did wow so which I don't think ever happened reacts over the course of YC yeah 3X over the course of YC which I
• 10:30 - 11:00 don't think has ever actually happened on average on average it was only the best companies that did that which is the top quartile or something right so the companies are better the general thing that is true is just that the time it's taking to reach $100 million in annual revenue is trending down yeah and not only that um we had dinner with Ben Horowitz recently and remember he was saying when they started Andries and horowits the common understanding was that in any given year
• 11:00 - 11:30 there'd only be 15 companies that year that would even make it to $100 million a year revenue and uh they said they ran the numbers the last 20 years and every decade the number of companies that could actually make it to $100 million went up by 10x so what was 15 per year maybe 20 years ago I mean we're talking about 1,500 companies a year that have a real shot at actually making that number and when you combine that with what
• 11:30 - 12:00 we're seeing in the summer and fall batches it's not that surprising and Jared had a really good argument on our last episode about how vertical AI is going to enable this to have thousand 500 plus companies to bloom yeah that's why it's growing so fast it's because the value propped to companies of these products is so incredibly strong that like they're just flying off the shelves cuz like I companies are smart and they
• 12:00 - 12:30 can do an Roi calculation and when the ROI is fantastic all these all these truisms that people believe about Enterprise Sal cycles and like quite a taste to get big Enterprise deals go out the window because companies are smart and they'll make rational decisions you know har there there's another way that this broken favor of startups that I was thinking about um it's hard to even remember now but a year ago one of the things that people said a lot was that these llms are not reliable enough to deploy in the Enterprise they
• 12:30 - 13:00 hallucinate that was why a lot of people said like the these pilots and PC's like won't translate into real contracts is because yeah it's too risky of a technology um for people to actually deploy deploy yeah and not only is it transl into real Revenue but translating into real deployments that are like being used at Large Scale you know doing thousands of tickets a day and I think it's because we've learned how to make the agents reliable via the kinds of techniques that Jake talked about when
• 13:00 - 13:30 he was here and just the all this infrastructure has grown up around around the bottles it's enabled people to make them reliable yeah that's actually a big Trend uh this year is this concept of uh thinking of AI more as agentic that is a term that kind of bubbled up a lot this year it was not in the bubbled space of conversation last year last year was more about a lot of things that were kind of very chat-like chat I mean that was kind of the Riff on it but now remixed into a bunch of
• 13:30 - 14:00 agents for XYZ and we just I mean you just put out Gary just put out a great explainer video about computer use from Claude but just the capability of the models keeps pushing in the direction of just being able to do like complex multistep things and actually take over your computer and call other applications and and perform complex tasks that just didn't seem possible a year ago what about regulation seems like we sort of dodged a bullet there with uh 1047 and uh it looks like some of the Biden EO uh is not that likely to
• 14:00 - 14:30 survive the Trump White House TBD what that means in the longer term but certainly one of the things that we were very worried about was that some certain amount of math Beyond a certain level would suddenly become illegal or require registration at your local office it's certainly been a weird time to be in Tech because I've never experienced um software and Technology intersecting with politics so much and in particular I'm not used to genuinely
• 14:30 - 15:00 caring about National politics affecting startups in a YC batch or just you know companies that are less than a year old but it did really it did really for a moment was worrying it wasn't clear whether startups would actually be able to build Innovative AI applications versus suffering from regularity capture from open Ai and a few big players we're obviously very glad it broke in the favor of startups seems like we're still in the early game right I mean I mean it's very easy to see that
• 15:00 - 15:30 um the platforms themselves really will or could possibly resemble you know the win32 Monopoly right Windows uh has access to the apis they in fact know all the stats about what's working on their platforms and guess what they can build it into their platform you know we sort of breathe a sigh of relief right now in 2024 but um you know it's anyone's game honestly like these things are moving so quickly I wouldn't totally breathe your
• 15:30 - 16:00 last sigh of relief yet you know it's we we got to keep working on this okay so it's clearly been a great year for startups um what else has been happening who else has it been um a great year for do we think there certainly been some big funding rounds right like open AI unsurprisingly has raised huge amounts of capital scale yeah even within YC though we've seen like scale AI has really broken out this year $6 billion for open AI 1 billion for uh scale a billion dollars for SSI
• 16:00 - 16:30 the new ilas sger startup scale I think is just worth talking about because it's such a classic startup Story I mean you were there in the early days right you interviewed them for YC um what um tell us what the idea was that they interviewed with and how they ended up landing on what you know is probably one of the best startup ideas of the last 10 years the fun thing about the scale. a story is that it is the sort of epitome of the like classic YC startup story
• 16:30 - 17:00 there's other kinds of startups that get started you know like SSI for example that's not a typical YC startup story where like some very wellestablished people raise a billion dollars with like a PowerPoint pitch but like scale. a is like the classic story of how like young programmers can just gradually build a like10 billion company over time by being like smarter and harder working than anybody else and so yeah when Alex uh interviewed at YC he wasn't working on anything related to AI it was a completely different idea and the idea
• 17:00 - 17:30 for scale.ai kind of got pulled out of him by the market and it's it's actually still like several pivots cuz like uh the the original idea at YC didn't have anything to do with AI and then for a long time he was basically doing data labeling for the self-driving car companies they applied as I remember they applied like a healthcare related idea yeah it was a website for booking doctor's appointments okay cool and then they pivoted during the batch um do you remember how they came up with the data labing Li cuz this supposed to be in was
• 17:30 - 18:00 this 2016 yeah the way they came up with the data labeling idea was that Alex had worked at quora and quora had to do some data labeling for like moderation and stuff and so at the time the Big Data labeling service was Amazon Mechanical Turk and they were deemed unbeatable because they were like run by Amazon and Amazon could throw infinite money at it it was always at scale it was at like like quite large scale already but Alex had a unique Insight which is he' actually used mechanical torque at quora and he knew that it kind of sucked to
• 18:00 - 18:30 actually use it and so he he he had this sort of like unique Insight on the world and so he just tried to build a better mechanical TK basically the version he would have wanted when he was at Kora and as I remember it like really the their early traction came almost entirely from One customer Cruise right yes which needed to do tons of data labeling on all the images that the cars were taking as they were driving around San Francisco you got to like draw a circle around the traffic light and thing like that and the the cool thing about scale is that they've actually caught two waves so they you know
• 18:30 - 19:00 accidentally caught the first wave of all these self-driving car companies because ml took off at that time in computer vision there was just an unprecedented demand for labeled data for training sets that just hadn't existed before and so they were able to ride that wave and then as that wave was like crusting llms got big and all of these companies needed to do rhf at very large scale and scale was just like perfectly positioned to move into that
• 19:00 - 19:30 business as well yeah like I think the scale story is are so interesting CU it was pre llm it was it was clear a multi-billion Dollar business anyway and llm it called The llm Wave which is now proped into probably it's going to be like a hundred billion doll plus company and I'm seeing that at the ground level too where many companies I had that maybe finished the batch even pre- the batch didn't have an idea pivoted into an AI idea that's taking off like I'm just seeing much more success in Founders who who waited out and can find
• 19:30 - 20:00 an idea that they just couldn't before I have a company from a year ago they Prov they proved the whole batch they couldn't find um a great idea it actually took them six months after the batch until they realized um one of their parents ran a dentist office so he just decided to go hang out at the office to see like if there was anything he could automate uh and they just ended up building an AI back office for dentist offices and now it's just like their week over week growth is fantastic it's doing really really well and I'm
• 20:00 - 20:30 seeing lots of cases like that spring up definitely seeing that as well I think there's something about the advantage of having all these very hardcore young technical Founders that are willing to kind of just bet the farm and go all in on just a little bit of a glimmer of oh this is where the future is going to be let me just try it and then it actually ends up working I just St with the dentist I have a lot of teams that pivoted as well into different spaces where they kind of found that glimmer like oh computer use came out and I have a couple companies that are working and
• 20:30 - 21:00 betting and going in that direction and it's like working well I mean it's still early I mean this is just a fall badge but that's cool too okay so what are some of the act the trends that we've seen what are some of the the specific Trends and waves that startups have been riding coming out of the batches voice AI is something we've talked about it's clearly um maybe the most promising vertical for AI right now in terms of just Ro traction do you think uh voice is a win or take all or will it be something that has sort of a 100
• 21:00 - 21:30 different verticals that are very tailored to those specific verticals that's literally one of the questions I get from some of our uh voice AI startups themselves they're like should I be going horizontal or should I just continue to grow with within my vertical it feels to me like voice itself vo is um I just like a it touches everything and there's so many different applications for it that um you can there's probably infinite
• 21:30 - 22:00 applications to build where voice is the interesting element of it I mean things that just spring off top of your head like language learning applications I'm sure there's not going to be just one really cool voice AI power language learning application probably going to be multiple of them remote work like um teleconferencing there probably like a whole other area where there's interesting things to do with voice Ai and even within customer support we highlighted a number of companies we talked about last time um company parel
• 22:00 - 22:30 cap. a yeah it it turns out that customer support is not really one vertical there's like many different flavors of customer support and they're like very different on the inside once you get into the details because I think there's very specific types of workflows you need to do per industry and that's to the point of why vertical AI agents are going to really flourish I mean same thing for Bo it's just very different workflows if you're building the I don't know the voice agent to do customer support for an airline very different than doing it for a bank very different than not doing it for a BB SAS company
• 22:30 - 23:00 Etc yeah I guess that question of is there going to be um pure horizontal integration is sort of like saying will there only be one website yeah or it'd be like saying like they're just going to be both it'll be horizontal infrastructure companies that do really well and vertical applications because it to say otherwise would be like saying oh like stripe Powers payments on the internet and it's also just going to have all the most valuable applications that accept payments on the internet it's just not how it works like there's
• 23:00 - 23:30 enough value at just being the horizontal infrastructure layer so I'm sure there'll be great voice AI companies that just make it really easy for you to build your own voice AI application while there'll also be hundreds of really valuable vertical apps what are what are what are other trends that we' seen besides uh voice we were talking about robotics earlier there certainly we we are certainly working with more Founders building robots this year than I think any year ever uh what's driving that I have a x app team that's called weave robotics
• 23:30 - 24:00 that they're going to try to ship a real robot in 2025 it costs about 65 $70,000 but that's actually what it costs to have the actuators and the safety needed to actually have it work in your home I think it's actually driven by this idea that um the llm itself can be sort of the consciousness of the robot like am I doing this thing that you know my owner needs me to do um you know how do I actually interact with
• 24:00 - 24:30 them and the other people in the household uh but it's funny because then The Voice language action model that might actually do a certain thing like fold laundry that's uh almost tool use inside of the broader llm Consciousness so I feel like that's one of the things that I'm excited to see you know will it really work and I think we're going to find out this year I guess the way I think about it robotics is basically half Ai and half Hardware half of the part of the equation is starting to work
• 24:30 - 25:00 well the hardware is still hard the hardware is still very expensive yeah there's still there's some evidence that uh being able to actually do laundry for instance like that might be one of the first things that gets shipped I think the dream case for startups is going to be that you can build just the the AI or the software piece of it and run it on commodity hardware and do really great things the opposite case would just be actually if the two thing like if you need to be good at the hardware and the software and they like coupled together and need to produce both and you would
• 25:00 - 25:30 expect Tesla to be the obvious like winner in the space and it remains to be seen I'm pretty optimistic we have multiple companies I feel they trying to be creative on how to run the models on commodity hardware for specific use cases it still feels early it it feels like the robotics hasn't quite hit its like chat GPT moment yet maybe the moment is self-driving cars have been working in San Francisco I don't think it's talked enough people who don't live in San Francisco or like often don't
• 25:30 - 26:00 realize the extent to which these are fully deployed in San Francisco and regular people are writing them every single day yep I saw Tony from door Das recently and he said he exclusively uses weo like everywhere I live in Palo Alo and have no option for it but I would love to it' be amazing I mean the wild thing is there are only a few thousand of these deployed right now in the entire world and Frisco yeah what about big flops for 2024 uh you I seem to remember that we uh started one of our
• 26:00 - 26:30 light cone episodes all wearing uh Apple Vision pros and quests and uh we have not talked about AR since Diana what happened uh it hasn't happened there's this moment for a lot of the hardware that needs to be a lot more lightweight like we need to get to this form factor but there's actually constraints with physics to fit all that Hardware in such a small form factor in order to have enough computer and the Optics to fit is
• 26:30 - 27:00 just super challenging and I think there's still more actual engineering and physics that needs to be discovered and that that's it I think the algorithms are there but it's just lots of really hard hardware and Optics problems it's a it's a tough chicken and egg problem because there's not enough Hardware in people's hands for it to be worth it for app developers to build apps and so there's not enough apps for people to want to buy the hardware and I feel like the people who did buy like the the killer application so far seems to be using it as um uh a really large
• 27:00 - 27:30 monitor um uh it and it does work very well for that yeah you've actually retained as a as a user Gary right yeah it's great for watching movies yeah maybe the one device that I think actually been playing and actually feels good is actually the meta Ray Barn it doesn't have any of the actual displays but I really like it for the audio and voice and one workflow I've been trying out is actually using the MAA Ray barn
• 27:30 - 28:00 and connect it to uh any of the voice modes for either chpt or Claude and kind of have a conversation with it about a topic oh I haven't tried that that's an interesting idea yeah yeah that's a great idea that is like a fun thing that I've been doing and just chatting with myself maybe look a little bit like a crazy person while you're walking but uh it's been fun to kind of learn about different topics should we talk about AI coding 2024 was the year that AI coding really broke out yep and we had the majority of the of YC Founders now use
• 28:00 - 28:30 cursor or other AI idees they just like exploded over the summer Devon proved that you could like fully automate like large programming tasks yeah all that was this year that was pretty wild replate agents continue to improve like I'm hear more anecdotal stories of people building replate apps on like their way home from work being really impressed like repet took this technology and popularized it among like non-technical people for the first time
• 28:30 - 29:00 that's really crazy and even more uh lower technical version is uh anthropics artifact where you can actually prototype very simple apps and chat with claw to build really FR simple front pages and then you can prototype stuff as a PM and show it to your engineering team and it's like a full-fledged working version yeah it's wild because it just means that you can one person can do so much more and do you think it's going to change the nature of how startups are actually
• 29:00 - 29:30 hiring are you seeing this yet like some of the founders I've met who recently raised their seed rounds coming out of YC um they're not really approaching it how maybe the classic advice would uh teach them in the P you know in the past you might say let me try to find you know more let me try to hire more people like I you know there are certain tasks that normally I have to find you know the person who did it at my competitor who did all of customer success and I
• 29:30 - 30:00 need to find that person who's under the person who runs that function and I've got to hire that person and promote them and uh they're going to come with all this knowledge and people networks some people are saying sort of the opposite which is I'm going to get my software Engineers to write more processes that use llms upfront and you know I probably will end up needing to hire that person but maybe after the series B or C and not right now yeah I think I've seen
• 30:00 - 30:30 that as well with companies after the batch where they're looking for engineers that have more upside and they're really fully native with a setup with a AI coding stack and part of the one of the clever interview checks I've seen is people do pair programming and watch them use the tools and you can really tell if someone really has Tinker with them it's actually an engineer cut that that is not only good at coding but also prompting and telling with when the
• 30:30 - 31:00 AI output is not correct I think the part of reading and evaluating the output of all these AI coding agents is actually a lot more critical yeah there's been an interesting controversy this past year about AI coding agents and programming interviews because AI coding agents basically broke the standard programming interviews that companies have been doing for years actually har I'm curious what you think about this since you ran a programming interview company I mean I guess the the interesting
• 31:00 - 31:30 debate is whether you should penalize or prevent people who are interviewing at your company from using cursor or one of these tools um to Ace your programming interview or whether you should just lean into it and adapt and uh test to see how productive they are um I generally think the way these things tend to or is more in that direction that I think it will just become you'll just be measured on your absolute output and the bar will go up I think like stripe for example were early
• 31:30 - 32:00 on this about a decade or so ago where they recognized that so much of what they needed their programmers to do would like build we web application and web software and not do hard CS problems and so the industry shifted away from the Google style interview of lots of computer science problems and whiteboarding to just give someone a laptop and make them build like a to-do app in like 4 hours so I think we'll just see the same thing happen where
• 32:00 - 32:30 people just the industry will just adjust and you'll just be interviewed using these tools and just be expected to do a lot more in like a 2hour interview than you are today to your point Gary around just the the startups like maybe how many people they need to hire or just like how how they they scale it seems too early to see like dramatic effects on that yet but one thing that I'm interested in is uh so watched an interview with Jeff Bezos recently and he said that well one he's back at Amazon working on AI and two
• 32:30 - 33:00 that apparently Amazon itself has like a 100 or maybe it was a thousand it was a surprisingly large number of internal llm powered applications presumably to just run Amazon the last time Amazon took something it ran for internal infrastructure and released it to the world was AWS which completely changed how startups are built so I'm curious to see if they have interesting applications to run Amazon internally that they'll just release out and suddenly like there'll be new Stacks to
• 33:00 - 33:30 just build and scale your companies on and we'll see the whole something that we've talked about in recent episodes of the 10 person the one person unicorn one of the applications they did talk about is they did this giant migration for a old version of programming language whenever you need to upgrade different versions of database Etc it's like a lot of work and they use llms for it it was like changing hundreds of thousands of lines of code and it would have taken engineering project of 6 months or more
• 33:30 - 34:00 it was done in weeks I mean Amazon's just such a perfect use case for like L empowered agents doing back office processes they must have just like just absolute goal buying of opportunities and they just launched their uh big foundation model actually that is uh starting to be top in some of the benchmarks as well so I think they're trying to be another Contender through this race that's interesting because uh like from the bottom up like certainly from some of the people who still work at Amazon maybe right out of college
• 34:00 - 34:30 many of them do not have access to llms or are actually barred from using it from in their day-to-day so you know maybe that's one of the downsides of organizations when they get big enough um you know the future is already here but it is not evenly distributed even within the same organization think but that boat sort of well for uh both open source and uh sort of self-hosting llms like I it's on my to-do list to build my own stack of Apple Minis and run llama
• 34:30 - 35:00 on my own little cluster on my desk I bought all the hardware to build my own machine but then we had a baby and it hasn't happened at some point I've been pretty excited in that uh you know YC has been operating back in person at San FR in San Francisco for some time but we got a real live demo day yeah all the way back so no more zoom demo days no more zoom alumni demo day you know we did alumni demo day right here in this uh office
• 35:00 - 35:30 right you know right downstairs that was awesome and then we took over the Masonic Center and 1,200 investors uh all in one room it was actually really great for the founders I thought because it was about uh a third as many Founders than the summer batch and it was more than 2x maybe 3x the number of investors that uh then who had came to our investor reception party so was like a ratio of 10 investors for one company
• 35:30 - 36:00 roughly so uh I think all of them had a really good time I'd almost forgotten how great the energy of an in-person demo day is like it's just not something that you can replicate over Zoom the YC demo days also always acted as the deao investor reunion in Silicon Valley because it's the one event that all the investors would reliably show up at and so they were really excited that we had brought it back because when when when we weren't doing it there was no
• 36:00 - 36:30 equivalent event sort of the homecoming for Silicon Valley yeah so now it's uh you know four times a year and it's the one time that uh all the top early stage investors in the world are going to come back to San Francisco for hopefully that week's festivities culminating in our demo day so it's a real celebration it feels like iners in general is back that's certainly another theme of 2024 certainly the late stage startups that we've been meeting with and speaking to this year one of the highest priority
• 36:30 - 37:00 items has been figuring out how to get everyone back into person back into the office think like the era of it's going to be remote forever is definitely gone I certainly good ridd yeah exactly right and then F like yeah in person is back and then San Francisco is back like a lot of thanks to you Gary the elections recently seem to have gone well like there's a lot of optimism I feel around San Francisco and and yeah we have a new mayor uh we're hoping that he does the right things and um you know we have a very thin moderate majority On
• 37:00 - 37:30 The Board of Supervisors uh but we did get rid of some of the worst people who created a doom Loop in San Francisco so I'm optimistic you know we didn't get everything we wanted but uh it's tracking in the right direction and I think as in startups as in politics you always you know way overestimate what you get you will get done in one year but you always way underestimate what's going to happen in 10 years I think it's going to take 10 years it's going to take 20 years but um just as startups
• 37:30 - 38:00 went from 15 companies a year to that could possibly make it to $100 million a year to 1,500 in any given year knock on wood um that you know I think San Francisco needs to be the beacon for all the smartest people in the world and that that's actually probably the thing that I'm hope most hopeful for is that we can actually just keep building so from all of us to all of you watching happy holidays and we'll see you in the new year
• 38:00 - 38:30 [Music]